Master Student of Urban Mobility Institute, Tongji University
I am currently in the third year year at Tongji University, pursuing a Master of Science in Urban Mobility within the Urban Mobility Institute of the College of Transportation Engineering. [Sep,2021-Sep,2024]
Before my journey at TJU, I laid my academic foundations at the College of Engineering of Nanjing Agriculture University, where I obtained my Bachelor’s degree of Engineering in Transportation. During my years there, I had the privilege of exploring the Route Planning Problems using deeplearning algorithms, which contributed to my bachelor’s thesis and earned a nomination for Best Thesis. [Sep,2017-Jul,2021]
At TJU, the Urban Mobility Institute is an interdisciplinary institute. With my background in Transportation, and under the guidance of several professors from the College of Surveying and Geo-informatics at Tongji University and the School of Resource and Environmental Sciences at Wuhan University, I was responsible for tackling and completing key research components of a critical R&D project. This experience led me to shift my focus to the field of high-definition maps, focusing on HD map modeling, dynamic information organization and management, and information interaction.
My research enthusiasm lies in applying AI to grounded tasks. My primary goal is to develop methods for efficiently extracting key information from crowdsourced big data to serve urban science. To achieve this goal, I identified three main challenges:
How can valuable information be effectively extracted from crowdsourced SVI to enhance the detection and classification of road features for autonomous vehicles.
What machine learning models and data fusion techniques can be developed to integrate diverse data sources, such as SVI and vehicle sensor data, for real-time updates.
How can the proposed methods for accuracy, reliability, and scalability in various driving environments.
The National Key R&D Program of China Leading by Prof. Wei Huang
In this project, we propose a High-Definition Map (HD Map) model, focusing on the dynamic information of HD map and its exchange format. We design a comprehensive content structure and data exchange format for the dynamic information of HD map and develop a Group Standard. Additionally, we propose an information interaction approach to be used between autonomous vehicles and high definition map for broadcasting and receiving dynamic information. The The phased research result has already been published in the Journal of Geomatics and Information Science of Wuhan University. Currently, our newest research is coming.
Undergraduate Thesis Research Plan Advised by Assoc. Prof. Yang Liu
We utilized a trained deep learning model to address the Route Planning problem, specifically focusing on how to efficiently plan delivery paths for vehicles with capacity constraints. Our enhanced algorithm is applied to vehicle path planning scenarios with 20, 50, and 100 nodes, respectively. The result indicates that trucks tend to prioritize nodes that are farther from the warehouse during the middle stages of the journey. Aside from the final route, which must cover the remaining nodes, the algorithm ensures that each route maximizes the number of visited nodes.